STARFISHES OF THE PHILIPPINE SEAS. 263 



The gist of this quotation is that Sladen did not adopt Perrier's 

 incorrect use of Dorigona. With such reasoning it is no great won- 

 der that the nomenclature of the Asteroidea became tangled. 



KEY TO THE SPECIES OP NTMPHASTER HEREIN DESCRIBED. 



o\ Marginal mouth spines (bordering on mouth of ambulacral furrow) 6 or 7, 

 exceptionally 8 (if 8, then rays not conspicuously slender, but if meas- 

 ured at proximal suture of the first pair of superomarginals which meet 

 medially, equaling length of first 3^ to 4^ superomarginals measured on 

 ambitus) ; some of the radial plates wider than long; mouth plates small, 

 and rays stout, the inferomarginals extending laterally beyond supero- 

 marginals. 

 6\ Mouth plates nearly or quite inclosed by the first pair of adambulacrals ; 

 numerous subambulacral spines ; abactinal granules mucronate. 



mucronatus, p. 269. 



6*. Mouth plates not inclosed; subambulacral spines few or none (granules 



only) ; abactinal granules at best only incipiently mucronate, and these 



few in number. 



c\ Five lon.gitudinal series of abactinal radial plates with the distal as 



well as proximal plates obviously wider than long ; no marginal or 



abactinal pedicellariae ; inferomarginals and superomarginals wider, 



the latter on rays, the former on disk; width of ray as measured in 



o\ equaling first 4 or 5 superomarginals euryplax, p. 264. 



c'. Radial plates not so obviously wider than long, the difference showing 

 on the proximal plates of the radial areas rather than on the distal ; 

 numerous abactinal and a few marginal pedicellariae; superomar- 

 ginals narrower on ray and inferomarginals narrower on disk; ray 



measured as in a\ equaling first 3i superomarginals dyscritus, p. 266. 



c'. Marginal mouth spines 8 to 15 (if the lowest number, or exceptionally less, 

 then rays also very slender throughout). 

 b\ Abactinal radial plates conspicuously wider than long (much as in mucro- 

 natus) ; rays very slender leptodomus, p. 272. 



b^. Abactinal radial plates roundish or hexagonal, not conspicuously broader 

 than long, 

 c*. Second superomarginals meeting in median line across ray (4 supero- 

 marginals in each intorbrachium) ; inferomarginals of ray very nar- 

 row and long, the length at middle of ray exceeding the height 

 (or thickness) of lateral face of ray; rays slender and delicate; oral 



spines, 10 or 9 atopiis. p. 285. 



c'. Fourth, fifth, and sixth superomarginals meeting in median line across 

 ray (sometimes third in young specimens), and therefore 6 to 12 

 superomarginals to each interbrachium) ; length of inferomarginals 

 at middle of ray equal to or less than height of lateral face of ray. 

 d\ Ray broader at base, measured as in a^ equaling first 4i siiperomar- 

 ginals ; supei'omarginals very broad, not tumid ; oral spines, 10 to 

 12 ; furrow spines 10 ; inferomarginal plates of interbrachia extend- 

 ing laterally beyond superomarginals moluccanns, p. 274. 



d". Ray narrower at base, measured as in a\ equaling first 3 to 4 supero- 

 marginals; if approaching rZ\ then superomarginals tumid; in in- 

 terbrachia superomarginal plates extending laterally beyond infero- 

 marginals. 



